COVID-19 Webinar Series Session 7 – From Data to Decisions: Evaluating State Capacity

12:00 p.m. – 12:05 p.m.     Welcome and Introductions

Sarah J. Dash, MPH, President and CEO, Alliance for Health Policy, @allhealthpolicy

12:05 p.m. – 12:30 p.m.    Moderated Discussion

Nirav Shah, M.D., MPH, Senior Scholar, Clinical Excellence Research Center, Stanford University; Collaborator, CoVidActNow, @niravrshah

David C. Radley, Ph.D., MPH, Senior Scientist, Tracking Health System Performance, The Commonwealth Fund, @commonwealthfnd

(This is an unedited transcript. For accurate quotes and presentations, please refer to the full-event video.)   1:01 Good afternoon. I am Sarah – president and CEO of the Alliance for Health policy. The alliance is a non-partisan resource for the policy Community dedicated to advancing knowledge and understanding of Health policy issues. Welcome to the 7th event in our covid-19 webinar series. We launched the series to provide insight into the status of the covid-19 response and shed light on remaining gaps in the system that must be addressed to limit the severity in the United States the alliance for her. 1:31 Policy gratefully acknowledges the National Institute for Healthcare Management foundation and the Commonwealth fund for supporting today’s webinar. You can join today’s conversation on Twitter using the hashtag all help live and follow us at all Health policy and we want you all to be active participants in the discussion. So please get your questions ready. I’m going to give you the quick technical notes on how you can ask them. You should see a dashboard on the right side of your brat web browser. Please select the speech bubble icon with the question mark to chat with us about any technical issues. 2:01 Issues, you might be experiencing and you can use that icon to submit questions that you have for the panelists at any time. We’ll collect them and address as many as we can during the broadcast. Finally. You can find the materials that accompany this webinar on our website along with a recording of today’s webinar. 2:17 The covid-19 pandemic is putting unprecedented pressure on the healthcare system and the economy today. Our panelists will explore how data and projections can help federal state and local leaders respond to covid-19 and prepare for Next Step first. We will hear from dr. Nirav Shah a senior scholar at Stanford University’s clinical Excellence research center near Rob’s a board certified Internal Medicine physician who works extensively in patient safety and digital Health innovation. 2:47 He is also a contributor to the covid act now project a tool that helps policymakers evaluate the effective strategies to reduce the spread of covid-19. Next. I’m pleased to introduce. Dr. David Bradley a senior scientist for the tracking Health System performance problem program at the Commonwealth fund. David is also a senior study director at westat a research firm supports the health system scorecard project. Thank you both for joining us. 3:13 And now I’m going to turn it over to dr. Nirav Shah for opening remarks. 3:18 Well, thank you so much Sarah. It’s really a pleasure to be here and thanks to everyone who’s joined the call. We build models to predict the future based on what we’ve seen in the past. And right now we’re working off of data from Spain Italy and other countries to come up what we think might happen in New York, Florida, Iowa and across the United States. We don’t need models to tell us that things are going to get worse and possibly much worse before things level off, but we do need signals in real time. 3:48 Tell us where to focus our attention next slide. 3:53 So about two weeks ago actually exactly two weeks ago a new website was launched at covid act now dot org founded by a bunch of technologists in Silicon Valley and around the country max Henderson Jonathan Christ Tompkins Eagle Kaufman and Zach Rosen brought together a community of all volunteers experts in their respective fields to answer the question what’s going on first in California, and then they quickly spread. 4:23 Across the country and now we’ve actually been able to do it for over 2,000 counties Across America Next slide, please. 4:34 Projections were initially designed to make decision makers understand how their actions or inactions lead to externalities. When will your hospital system be overloaded. 4:49 If you do not act now that was the question being asked and if you click on a state next slide The point was to get everyone to stay at home initially by flattening the curve by providing up-to-date information with the best available data already. All of this model is fully available online next slide, please. 5:13 And it’s gotten a tremendous response. We know that officials in Michigan, Colorado Alaska Hawaii and other places have used this data in real time to make their decisions to enact stay at home or shelter-in-place orders. Next slide. This is an example of what the website shows and if you type in a county it’ll give you a county level information. 5:39 If you type in a state it’ll give you a state level information and what this shows is based on actual Hospital capacity in a given region given the number of cases either hospitalized or people who have died from covid or any other data that’s available the best available evidence from across the internet. What is the likelihood that your hospital system will be overloaded and if so, when relative to different actions, you could take to flatten the curve in this example, you can see that by the end of April hospitals will be overloaded in in one state. 6:16 However if they do social Think they can extend that period till end of June or beyond the reality is models are good at predicting the near future. They’re not as good as predicting the far future but short of any other information. This model has been used widely by the military and and many others to help identify opportunities and focus resources next slide. 6:45 Other models help us in different ways. I am an adviser to this group full disclosure Health whether dot U s– is powered by kinsa. They have a million smart. Thermometers spread across America in people’s homes that are connected to the internet they Bluetooth to your phone and what this does is it gives you a signal of where is fever occurring Across America. They have a predictive model that predicts flu accurately. 7:14 Out for twenty-plus weeks one, you know what’s expected for flew in a given City because you have a fingerprint of how flu looks like and works in that City. You can subtract out regular flu and come up with unexpected febrile Peaks. So this is people pulling thermometers out of their medicine cabinet taking their temperature and in real-time updated daily telling us where is something going on that’s not expected. 7:44 By seasonal flu in the community. Our best estimate is that many of these are actually covid hot spots. You can see certainly the lighting up in the Seattle and Washington area New York parts of Texas and Florida. This is also a very good tool if you read the New York Times from last week to track if your shelter in place is having an effect as the fever’s go down. Are you doing a good job of it? 8:12 Next slide please. 8:15 This is an example from their data that shows Santa Clara versus Miami-Dade and various interventions and how those interventions had an effect on the fever curve the a typical illness curve in those communities currently. This has a county level resolution. We’re looking to work to scale that up with more thermometers deployed to the city or suburb City level as well. 8:39 Next slide other models developed by other groups are using very sources of data to answer some of our most pressing questions is your social distancing working group called unicast has this social distancing scorecard on their website where they’re using real-time GPS signals from people’s phones to track how people are moving around in their homes in their communities relative to historical Trends so you can see in States like Hawaii and Ticket people are moving around a lot less than historically based on tracking of their GPS. This is all anonymised. You know it privacy protected. Whereas there’s room to improve in some of the orange States. 9:31 With that, I’ll conclude my remarks and leave more time for the questions. Thank you. 9:37 Thank you so much here of I want to ask you before we move to David. I’m just if you could take a moment to comment on a couple of just definitional questions. We’ve heard terms like they at home shelter-in-place mandatory shelter in place, you know, and I’m wondering if you can kind of comment on how the Kovac now model or other models that you’ve seen look at the assumptions of some of the underlying policy decisions in in terms of feeding into them. 10:07 All and you know, are there significant differences there? Yes, when we say stay at home, that means something different than shelter in place. The reality is that everyone is using the terms differently the New York Times has a very good summary of what states and counties are doing and and the definitions of what those terms mean. I would refer people to that for a longer Exposition. What we do know is that the more stricter policy, you know, ultimately getting down to the woohoo. 10:37 Han style lockdown where people are segregated those who have symptoms or are suspected or are positive away from even family members who are negative that level of lockdown leads to the quickest results. 10:52 What we know is that the faster you act the more you act the more decisively you act the better it is for us to recover and our economy to get started once again, Okay. Thank you, and I’m sure there will be more questions. Just a reminder to the audience. If you have questions, please go ahead and use that chat function in your webinar interface and feel free to ask them at any time. So now I’m pleased to turn it over to David Radley to talk about his research. Thanks. Go ahead David. Thanks Sarah next slide, please. So we’ll talk a little bit today about estate. 11:33 Eight Level Health System capacity to deal with the covid-19 pandemic the data that I’ll be presenting today was published in a brief by the Commonwealth fund last week, and I did want to just acknowledge my co-authors. All kind of will assign staff Jesse Baumgartner Sarah Collins Eric Schneider and Monday Abrams, you can download the brief that that that I’ll be talking from today at our at our website Come O fun dot org next slide, please. 12:02 I think we all know why we’re here today, but let’s just start with some stage setting we know the virus is spreading quickly. Just last week on the 26th seven days ago. There are only a handful of states where the number there was more than 25 confirmed cases per hundred thousand State residents adult State residents by by as of yesterday is Count 33 States had at least 25 cases per hundred thousand adults. 12:31 Eight and twelve States and climbed up over that 75 case per hundred thousand adults New York, which we all know is a bit of an outlier here as of yesterday’s data. I think was about 550 cases per hundred thousand adults in New York state and that’s by far the highest now to be clear. This rapid increase does reflect both an increased in testing capacity and the national spread of the disease of the disease, but even though testing capacity is increasing. 13:01 there’s still not enough tests being performed and we know we’re seeing news reports of major backlogs and test processing that that ultimately mean that even these counts likely under represent the actual prevalence of covid-19 in populations across the country next slide, please so as the virus spreads State leaders are really left trying to mobilize their Healthcare resources in there quickly coming to terms with their capacity to be able to handle potential serve. It surges in covid-19 cases. One of the things we do at the Commonwealth fund is we try to catalog different aspects of the health care delivery system. So we pull together state level estimates on population risk Hospital capacity and physician Supply in hopes that we could give context to state and federal. 13:53 Policymakers we’re working through their own response scenarios. I’ll walk through specific data points over the next few slides, but I do want to draw your attention to the Excel file that you can download from our website that has all of the state-level data that I’ll be discussing today. You can download all of that data from our website in the in that data set that you could download simple Excel file. 14:15 It even has some information about health care Access healthcare insurance coverage and affordability concerns that sir, Certainly will matter as States develop responses to covid-19, but that up, but that I’m not necessarily talking about in this presentation today next slide, please. 14:33 So even though we’re all equally susceptible to to Contracting covid-19. If we’re exposed to the underlying coronavirus. We also know that older adults and those with certain chronic illnesses have the highest risk for developing the most severe symptoms this map highlights the share of individuals in each state who are either over age 65 or who are between the ages of 18 and 64 with either chronic. 15:03 Structure of pulmonary disease asthma diabetes or who are severely overweight with a BMI over 40. There’s not huge differences across States and you know, every state has a significant high-risk population, but that said there is a cluster of States particularly in the southern part of the country where there does seem to be a higher share of people in these high-risk categories that may be particularly susceptible if indeed the virus spreads to those areas. 15:32 next slide so here we’re looking at how States differ on each of four measures of physician Supply. We look at physician capacity and ambulatory settings. 15:44 These are mostly Primary Care Medical Specialists who might be expected to play a role in triaging sick patients to either to who may need to go to a hospital for more intensive care or you know performing some level of care management for those who may be less acutely ill the We also look at the current capacity of hospital-based Physicians including those in general Practice Medical and surgical surgical Specialties and emergency medicine who may be more accustomed to treating and managing patients and more intensive care settings. And then finally we look at the added capacity that may be available in a surge scenario and I’ll tell in the next slide. 16:28 I’ll talk more about what we mean by a search in Aereo, but just to help interpret what’s happening on this Slide, the the lighter God’s relatively higher Supply in the darker dots represent relatively lower Supply. 16:40 And what we see when we look at the underlying data is that there’s actually a fair bit of sort of natural variation across States in position Supply. 16:48 I think that’s really to be expected think by itself, you know variations supply of Physicians by isn’t necessarily an indication of how good or bad or how prepared or ill prepared a particular say it is, you know rather it’s a marker of what resources may be available for mobilization when the need arises Slide 6, please. 17:12 So here I did want to just spend a minute explaining what we mean by surge capacity. This may be a concept that that may be new to some of you. So in our analysis we distinguish between Physicians who self-identify as being primarily responsible for performing patient care as their sort of main daily job from from those Physicians whose whose responsibilities are not patient care. 17:38 Maybe they’re serving in a City setting and they primarily teach or they primarily do research they could be administrators, you know that kind of role they’re licensed or at least have had medical training. I actually don’t know whether or not they’re licensed but we do know that they’re meant that they have medical training but they’re not serving in a patient care role on a daily basis Under The Surge scenario. 18:04 We assume that those those physicians in that latter category would actually To providing Frontline care. So using New York as an example that translates to about an extra 20 100 Physicians that could potentially be available to help you, you know as sort of the need increases where we need Physicians to care for patients next slide slide seven, please. 18:32 So on slide 7 this is set up similar to the to the to the slide. We showed you two slides ago that looked at physician capacity. But this one’s set up to look at hospital bed capacity per tick in particular the capacity of acute care acute care hospital bed Supply intensive care hospital bed Supply and ventilator Supply again. 18:55 We see a fair bit of variation across States if you were to look look at that spreadsheet that I referred you to the And then you could actually see the rates per hundred thousand there. So there’s a fair bit of variation across States and the amount of sort of hospital bed capacity that they have Again by itself that reflects sort of a natural variation. It reflects what the health care markets sort of can sustain in normal times and it isn’t necessarily again an indicator of how prepared are ill-prepared possible systems in a particular state are rather. It’s an indication of What kinds of resources may be available? 19:32 Liable as Governors and state and local Healthcare leaders are trying to mobilize resources. 19:38 Slide 8 please as we’re looking across both position Supply and hospital bed Supply. The one thing that’s clear is that that that all states are going to face big challenges regardless of what their level of Supply is today? 19:57 All states are going to face big challenges should covid-19 cases surge in the coming weeks and months, you know states have really emerged as the Locus of policy attention and policy action to try to mitigate the virus has spread like dr. Shaw said, you know states of the what are the sort of the policy entities that are enacting stay at home and shelter in place orders states are mobilizing Healthcare resources as best they can across the state and and it in you know, it’s so much of this of this response really is coming down to what state leaders are able or able to do. 20:39 Um with that I’ll turn it back to Sarah and and I think open it up for questions. Thank you. Great. Thank you so much both of you and I we have about 10 minutes for questions. I want to ask first just in keeping to your last slide David. I mean this question about surge capacity versus the maximum capacity of a healthcare system. 21:00 And you know, have we even seen Healthcare systems working at the the the maximum capacity that’s measured in your models and maybe who comment on on that note. I look for Europe to comment as well. 21:13 It’s tough to know for sure because you know, the the data that we’re able to collect here really is a couple of years old. Honestly, I think New York is probably operating Beyond Where what our search scenario is, you know, we saw last week maybe as earlier this week. It’s all running together. I’ve lost track but you know Governor Cuomo has his literally asked for for healthcare providers to come into our state from other states because the need is so great. 21:41 We in the New York City area. So honestly, I it’s hard to know because we don’t know how many, you know sort of in real time today how many Physicians are actually providing care to patients in hospitals and the data that we have the capacity that we have actually represents sort of the current state of the world in 2019. But by all accounts it certainly seems like New York at least is operating Beyond, you know, our search scenario. 22:10 Yeah when I was Commissioner of Health New York during Hurricane Sandy. We operated hospitals in the New York City area at a hundred and eight percent capacity for almost three months because of the large hospitals that shut down that was barely sustainable on staff on everyone that the taxing resources Etc right now Montefiore is currently above surge capacity and there are many others but there are very few places that can do things like New York where they take an entire Hospital the hospital. 22:41 Special surgery shut it down and turn it into a covid hospital. So to the extent that New York is unique in some ways. I think we have to be very concerned about what surge means especially in large parts of Rural America where they don’t have ICU beds. 22:58 Thank you. And can you give us a little bit of a sense of what it looks like on the ground that you were sharing that you’re you’re you’re back in practice right now. Would you be able to just share a little bit about that perspective for so I’m practicing at the Stanford Express Care doing video visits and triaging patients with covid. I’ve had many patients over the past week who are clearly covid positive, but because of the limited amount of testing available, we are asking them to stay at home. 23:26 Home and trying to self quarantine them to the best of their ability that is a policy that has not worked in Italy or in China or elsewhere and yet because we don’t have sufficient testing we aren’t able to actually test everyone who we believe has covid the rolling two-week promise that we’ll have enough tests continues to roll. I am hopeful that this time it will we will have in two weeks sufficient capacity. 23:56 Only after we start to actually understand how much covid is in the community. Can we make a difference the reality in all of the numbers those estimates of covid are based on absolutely inadequate testing the best numbers we have are related to covid related mortality, but that’s a late indicator short of that. 24:17 The best number that we should get today from every state in America is the number of people hospitalized with covid about half of those states are reporting that we need to change that right away. 24:31 Well that leads into a great question from somebody in the audience, which is what additional modeling would help Public Health officials and Health Systems to manage this crisis that you haven’t seen developed or published yet. So what unanswered questions remain that would would or could be estimated by modeling? 24:50 David you want to start. 24:54 I was outside your the modeling expert public health expert but I can tell you from a policy perspective. We’re trying to figure out what we need where we need to identify the resources and concentrate them in real time. The surges are happening one state after another we understand the needs of from these basic models in terms of based on hospitalized patients. 25:18 What we haven’t yet seen our good data on how effective are These strategies are in tamping down and flattening the curve that relative mix across rural and urban areas that relative ability based on different classifications of what counts as an essential business. All of those are needed by policymakers now so they can say, you know, this business is essential or it’s non-essential these policies work or don’t work today. 25:46 We have grocery stores as essential businesses, which of course they should be but they can also serve as Central Herbs where the virus is most quickly spread if we don’t have the kind of policies in place that other countries have put in around people visiting grocery stores. So there’s many questions. I’ll leave some in the comments afterwards. 26:07 Thank you, and I just want to have one. 26:12 How’s it going? Say, this is David one edition that I just wanted to make on top of that is is you know Healthcare happens locally and a lot of the data that we’ve seen it. It’s just starting to sort of get down lower than a stay you. Dr. 26:27 Shaw’s covid act now came out with kind of level data as of yesterday, but but but the and there and there are some sort of case counts and a few a few measures of death counts at the county level, but even counties are actually bigger than And where this thing is, you know where people get care and to the extent that we can understand and use modeling to do it to get down to smaller and smaller levels of geography. I think I think that’s an opportunity to to use that kind of technology to help us understand the spread of the virus better, too. 27:03 Thanks. Well, that’s a that’s a great segue. Actually, we got a question from the audience, you know, basically saying it’s great that the government is recommending stay at home. Of course, we know that very kind of from state to state but they’re noting that often our communities don’t have the resources for populations to properly stay at home and asking if there are any indicators that we can use to measure how well a population can actually adhere to these social distance and guidelines. So if you could comment on that and then here I want to ask you what specifically what policies have other countries. 27:33 In place for visiting grocery stores, or there are some examples that you can share. 27:38 Sure, so to the extent that I gave an example of one way of looking at how much Mobility is being adhered to or lack of Mobility with the unicast data. There are many other examples of trying to get real-time Mobility data to understand whether a policy is working or not and they’re getting to the finer and finer level and they’re doing so by while still protecting individual privacy, which is what’s also very important for us the part of the problem that I wanted to address. 28:08 So is that there’s a one of the most pressing questions is how soon can America get back to work last week and the week in the last two weeks. We’ve had 10 million people file for unemployment unemployment. America is insured through your job health insurance come through your job. If you don’t have insurance, how are you going to get health care that has not been answered and in states that haven’t expanded medicaid. That’s an even bigger problem. 28:37 So to the extent that there are a lot of questions that we need to answer. I think it’s a false choice between picking lives versus livelihoods. We need to have both we need to Tamp down people keep them at home and hence social distancing other countries including India have had a folks lineup for grocery stores by certain timings of the day. 29:02 You’re only allowed to go at a certain period of time you try to get folks who are under 30 to go you To have Central delivery of food as opposed to people going to get their own groceries. All of these are very hard to do at scale. But we do need to figure out what we can do about our Central hubs that also are not completely Tamp down yet. 29:26 All right. Well, we are getting we are running short on time. I just want to ask one last one last question before I get to my final question. So can you just share a little bit about what we know today about the demographics of covid-19. And you know, if you could share that and you know, are there other settings we should be looking at like the correctional system as well as we look to this crisis. 29:50 Yeah, the demographics in the US are a little different than we’ve seen elsewhere. We have a younger shift toward people who are getting covid. We are still seeing that children under the age of nine are not dying of this disease. It’s not that they’re not getting it. But the younger you are it seems the more protected you are but the reality the latest data also showed that about half of the people who are ending up on a ventilator don’t make it. So we need to be concerned. We need to act now and we need to flatten the Curve. 30:20 avoid death Alright, let me ask you to view just in the next 30 seconds. What should we be looking for in the next week as this situation evolved? 30:34 There are David maybe I’ll start go ahead. 30:41 David go ahead. This is the this David so in the next week, I mean, I think what we’ll be looking for, you know, we’re going to continue to monitor the case counts. We’re going to you know, and Johns Hopkins has those data as as you know, with caution will be looking at those because of course the the testing concerns that both in the Rob and I brought up but we’ll be looking at the case counts will be looking at the death counts. 31:07 I think we’re going to be looking at some of the states that That we haven’t seen huge surges in yet, but that sort of profile as you know, potentially at risk States in Florida is one where it’s a little bit older population with the chronic disease data that I showed you. They’ve got a little bit higher prevalence of the chronic diseases that make people particularly since I’ve susceptible to more severe symptoms and they were later to the game to enact a Statewide stay-at-home order. 31:39 And so I think for me that’s that’s where I’m gonna be paying attention. 31:45 I think that as the data points out. It’s a crisis in every state of the country over the next week. People will come to that realization. They will understand that this is something that we collectively must do by individual actions of staying at home. And I hope that as we start to see use of other testing serologic testing, we will start to understand the true picture of covid in the community, which I suspect is going to be exponentially greater than we think right now. 32:14 Now simply because we haven’t done the testing. 32:18 Well, thank you both. We will also be watching closely with you. So we are unfortunately out of time, but I want to thank you so much for joining us. Thanks to our audience. You will be able to find a recording of This webinar on our website following this afternoon as long along with the recording in case you missed it. We also had an exclusive interview with dr. Rebecca caps the director of the center for Global Health Science and security at Georgetown University, which you can also find on our website and please join us next week. 32:48 A conversation with dr. Mark McClellan about testing treatment and vaccines. So thank you so much again to nirav Shah to David Radley. Thanks to a commonwealth fund and the Nixon foundation for their support of this webinar. Thank you to our audience for joining us. Be safe wash your hands and take good care. This concludes the webinar. Thank you.